Chapter 5: Some Aspects of Hartshorne’s Treatment of Anselm by John E. Smith

Existence and Actuality: Conversations with Charles Hartshorne
by John B. Cobb, Jr. and Franklin I. Gamwell (eds.)

Chapter 5: Some Aspects of Hartshorne’s Treatment of Anselm by John E. Smith

John E. Smith is Clark Professor of Philosophy, Yale University

This symposium devoted to critical discussion of the thought of Charles Hartshorne provides an opportunity for me to press further some points I raised earlier in reviews of two of his books, The Logic of Perfection (1962) and Anselm’s Discovery (1965), both of which had to do with the original ontological argument and the import of Anselm’s meditations. I confess that I have always found myself ambivalent as regards that famous argument. That is to say, I am not confident in giving an unambiguous answer to the question, Is the Argument valid? By contrast, for example, I do not hesitate to say that Hume’s claim "whatever is distinguishable is separable," is wrong, but in the face of the ontological argument, I hesitate. On the one hand, the argument, understood in its proper setting, is not just so much nonsense or empty verbiage, for it contains a crucial logical transition pointing to a necessary relation between concepts, which, at the very least, can be argued about. And, as Hartshorne has pointed out, many of the "refutations’’ of the argument have been based on faulty apprehensions of its meaning or upon dogmatic assumptions such as the thesis that no existence can be derived from mere ideas or that existence is not a predicate. On the other hand, like Royce’s argument for the absolute from the possibility of error, one has the sense that the ontological argument establishes something, except that it is difficult to say exactly what that something is. It has always seemed to me that, on any interpretation of the argument, it enjoys an element of superiority over the cosmological arguments in its starting with the idea of God and of perfection rather than with an other. For the approach to God through the other must be limited by the fact that one reaches only so much, so to speak, of God as can be manifested in the nature of that other.

Generally speaking, I regard Hartshorne’s treatment of the argument and his tracing of its subsequent history as making at least three distinctive contributions to this philosophico-theological topic. There was first his return to the original argument without, we may say, benefit of Descartes, who confused the issue by asking for the cause of the idea of God. This return was accompanied by a new emphasis on the importance of the nature of God achieved by a reinterpretation of the "that than which nothing greater . . ." formula. The question of the divine nature was often thrust into the background because of exclusive concern with the divine existence. Hartshorne rightly redressed the balance in calling attention to the abstract character of existence taken by itself. Second, Hartshorne, with the aid of his neoclassical metaphysics, was able to show the need for real modes of Being in which the distinctive type of Being in question makes a real difference, especially in relation to divinity, in opposition to Kant, for instance, who thoroughly disconnected modality from the content of the concept, and thus found himself left with but one mode, that of spatio-temporal existence, as the matter for knowledge. Third, Hartshorne’s critical review of the assessments made by later philosophers of Anselm’s argument opens the way for a reversal of the usual procedure. Instead of attending only to a given philosopher’s verdict on the validity or invalidity of the argument, we are led to inquire into the validity of the standpoint from which the judgment was made. In short, the ontological argument can be made to stand as a touchstone for philosophical positions; what does the verdict of a given philosophy on the argument tell us about the assumptions and the viability of that philosophy itself? This reversal, in which Anselm and his followers no longer automatically stand in the dock but instead assume the role of prosecuting attorney, provides a new perspective on the entire discussion of the ontological argument, a perspective badly needed in view of the sorry record set by a host of past philosophers who, for the most part, failed to penetrate the substance of the argument because they already knew by heart the litany whose refrain is "Existence is not a predicate."

Turning now to matters of detail, I would like to press the point I raised in my review of The Logic of Perfection concerning the status of logic; my contention is that unless, as Hegel, Peirce and others have held, logic has an ontological reach especially as regards the modal categories the sort of reasoning represented by the argument must fail. This holds true for other metaphysical doctrines, including the basic one set forth by Hartshorne in his paper "Some Empty Though Important Truths.’’ The underlying problem concerns the status accorded formal logic, especially since it has assumed symbolic, mathematical form. Does it reflect the nature of reality, or is it a merely formal structure governing the use of language’? In short, are we to have no more than "logic without ontology’’? I believe that Hartshorne takes too lightly the force of the view according to which logic marks out the domain of the "necessary" -- sometimes construed as the tautological -- while the "real" coincides with contingent existence in the domain of fact. The consequence of this juxtaposition is that the "real’’ and the necessary are mutually exclusive.

Hartshorne appears to accept the formal/factual dichotomy,1 thinking of God as belonging to the domain of necessary truth rather than to the side of fact. I do not, however, understand how Hartshorne’s philosophical theology can succeed unless based on a logic with ontological import. His principle of ‘‘modal coincidence’’ is, I presume, intended to resolve the problem, but if it does, it is only because the modes are real and not only logical or linguistic. Hartshorne suggests that Carnap’s "meaning postulates" allow for the introduction of analytic judgments other than those that are merely logical, and presumably assertions about God belong to this class. He goes on to say that "it may be" that Carnap’s proposal is the key to reconciling the logical meaning of necessity with the ontological. This seems to me the central matter and one cannot take it too lightly especially in view of the fact that, in Carnap’s treatment of modalities, only the modal property ‘‘contingent" correlates with the "factual" taken as a semantical property; all other modal properties correlate merely with L-forms.

I do not overlook the fact that there has been much discussion in the intervening years -- the shaking of the foundations of the analytic-synthetic distinction, for example -- which surely does not leave the situation unchanged. As regards the dominant climate of philosophical opinion, however, despite all the disclaimers that have been made concerning classical empiricism and positivism, the old dichotomy between a domain of sensible fact on the one hand, and sets of logical forms on the other, seems to persist and behind it the old dogma that where we have necessity we have merely tautology, and where we have fact or "experience" there is no necessity. The reason I addressed the problem to Hartshorne initially is that, from one end of his thought to the other, he has made strong claims in behalf of his use, presumably in contrast to that of some other speculative philosophers, of modern formal logic, and I wanted to assure myself that he was under no illusion concerning the status accorded formal logic by many logicians and the force of the attempts to have, in Ernest Nagel’s expression, "logic without ontology.’’

I can express my point through an example, almost certain to be unfamiliar to most, taken from Royce, a thinker studied by both Hartshorne and myself, and, I should add, a philosopher who had his own somewhat transcendentally colored ontological argument for God. In 1908, at the International Congress of Philosophy in Heidelberg, Royce delivered a paper entitled "The Problem of Truth in the Light of Recent Discussion." The interesting and quite surprising substance of this paper is an enthusiastic endorsement of the then new mathematical logic and studies in the logic of mathematics associated with such thinkers as Russell, Frege and Dedekind. Royce fully accepted Russell’s logicist thesis -- the voluntaristic twist given to it by Royce we need not consider here -- that there are "absolutely true propositions" in pure mathematics which are, in turn, based on absolute truths of pure logic. The point germane to this discussion is that Royce was not only heralding the new logic for its concept of truth, which he regarded as far superior to that of pragmatism, but he was also assuming that this same logic could be used for the development of an exact metaphysics of the sort exemplified in the Supplementary Essay to The World and the Individual. Royce did not live to see the fulfillment of his high hopes and the development of this logic by some logicians not only into an instrument for the elimination of metaphysics but into a formalism and conventionalism in which truth in Royce’s sense no longer figured. Since Hartshorne’s knowledge of the developments that have taken place in logic over the past few decades is far superior to my own, I would be most interested in having his opinion about the general issue of the status of logic vis-à-vis metaphysical argument, especially in relation to recent discussion. And, in inviting him to respond to this query, I am not unaware, as Hartshorne himself has noted, that we must not take it for granted that the house of logic is in simple and good order, another indication, in my view, of the impossibility of disconnecting purportedly formal instruments from basic philosophical issues.

To begin with, I believe as I indicated earlier, that Hartshorne has done a great service to the odyssey of the ontological argument in the careful way in which he returned to the original text with its quite remarkable combination of meditative experience and rational articulation, as the basis for assessing the treatment accorded Anselm’s reasoning by subsequent thinkers. I would reaffirm my agreement with Hartshorne on the absolutely essential point that the ontological argument, properly understood, asserts that God’s existence is either necessary or impossible and, since there are no other alternatives, the argument cannot be discussed as if it involved merely the alternative of existence or nonexistence. This mistake has been the one most frequently made, and it finds its roots in Gaunilo’s example of the island; this line of thought received further support from the nominalist strain in modern philosophy wherein all real modes were denied with the exception of sensible existence. I believe that this criticism holds quite apart from the resolution of the question whether it is legitimate to speak of the presence of two arguments in the Proslogium. The reason is clear: the discussion about existence as a perfection, as if that were all that is involved, does not make explicit the far more important point that, in the case of "God," properly understood, nonexistence was never a real possibility, a consideration entirely overlooked by those who blithely say that, of course, we all know that no existence can be derived from "mere" ideas. The latter point brings us to what I take to be a novel and illuminating idea in Anselm’s Discovery, namely, Hartshorne’s answer to the question of exactly what Anselm discovered.

According to Hartshorne, Anselm was engaged in a meditative analysis of what it means to believe in God from within, as it were, since the believer is involved in a self-examination. From this starting point, Anselm is said to have shown that if believers understand their faith, they "are the only ones who do understand it" (p. 22), from which it would follow that it is only lack of understanding which leads a person to reject theistic belief. The positivist, according to this account, can avoid this conclusion only if he can consistently make good the claim that the term "God" is meaningless, which is a way of saying that God’s purported existence is not a logical possibility in any sense. In addition, Hartshorne claims that one who denies the existence of God explicitly cannot avoid Anselm’s conclusion under any circumstances, since his finding meaning in the central religious question at once prevents him from denying the necessity of the affirmative answer. On his own terms, Anselm is said to have shown that as long as "the fool" continues to conceive God he cannot consistently withhold assent to the necessity of the divine existence, unless he is using the term "God" in a sense different from what the self-understanding believer means by the term.

This, I believe, is an accurate representation of what Anselm intended by his meditation on the grounds of faith seeking understanding, and Hartshorne’s account clearly expresses the situation of the believer in relation to the two opponents. There is a question, however, about what limitation Hartshorne believes is imposed by the initial dependence of the argument on the idea of God derived from the faith and the self-understanding of the believer. In short, exactly what role is played by the faith from which the argument sets out? This question is not easy to answer, and it must arise ever again in any attempt to explain the approach through "faith seeking understanding." Hartshorne claims that, since all proofs have premises, Anselm’s argument must be based on the assumptions that faith is a real possibility and that the idea of God is free of inconsistencies. But these assumptions ("meaning-postulates") do not coincide with ‘‘faith" in the sense of the fides that stands in need of understanding. Meaning-postulates may indeed be required, but they do not furnish the appropriate religious meaning with which Anselm began the argument. I would agree with Hartshorne when he says that the argument is more subtle than the derivation of God’s existence from the premise supplied by the initial faith that God exists. If this is so, then the question arises as to what meaning is to be attached to the term "God" and how this meaning is to be circumscribed by faith. Moreover, did Anselm propose in the Proslogium, as he did in the case of the Incarnation in Cur Deus Homo, to demonstrate a doctrine to anyone regardless of their assumptions, or is the demonstration of God’s existence directed only to those "believers" who understand what is meant by "God" in a certain way? If the latter, how shall we know when we have started with the premises actually meant by a "believer"? Hartshorne does not pay sufficient attention to this problem because, in my view, he does not take seriously enough the historical, Judeo-Christian content underlying Anselm’s formula. Hartshorne does ask why Anselm chose his formula, "That than which nothing greater . . ." and answers, "I suppose because he takes it for granted that by ‘God’ is meant the universal object of worship, and if God could have a superior, then only the ignorant or superstitious would worship Him" (p. 26). Does this mean that Anselm was reflecting what "believers" mean, any believers, or rather what a "rational" believer must mean if he is really to talk about God? I find Hartshorne’s answer somewhat curious in that it envisages Anselm as having in his possession some generic category called ‘the universal object of worship’’ which determined the formula at the heart of the argument. While I would insist that "God’’ expresses a concept and is not only a proper name within a certain historical tradition, I believe that Hartshorne pays insufficient attention to the force of that tradition in shaping Anselm’s meditation. The "that than which nothing greater . . ." formula is Anselm’s attempt to express the perfection, majesty, and transcendence associated with the thought of God throughout the fabric of biblical religion -- "Thou shalt have no other gods before me."

The term "God" was obviously in use in the Christian communities long before Anselm commenced his meditations. While he couched the meaning of deity in his own formula, it is scarcely imaginable that he meant to do anything else but express what Christians mean by "God." Hartshorne admits that Anselm may not have succeeded in expressing the Christian understanding accurately. Unless, however, there were a faith content for "God’’ existing prior to Anselm’s formula, how would Hartshorne (or anyone else) know that interpreting "none greater can be conceived" to mean the "Perfect which cannot change" is "merely Greek" doctrine and not faith? Hartshorne in fact claims (p. 29) that Anselm adopted the Greek view and consequently that he sacrificed his right to say that his formula expresses faith. Such a claim, however, could be made only on the supposition that for "faith" or for the "believer" there is a meaning for the term "God" which Anselm may not have articulated correctly. The question then is, What is this meaning and how do we have access to it?

Since I would still urge the same line of argument on this point set forth in my review of Anselm’s Discovery, I shall quote one paragraph from that review.

The point is of the utmost importance because it concerns the meaning which the term ‘‘God’’ has for "faith’’ prior to its identification with Anselm’s formula. And indeed the term must have such a prior meaning if (a) we can significantly discuss, as Hartshorne claims, whether Anselm’s formula does express ‘‘Greek doctrine’’ or faith and (b) we can decide whether Anselm’s formula adequately expresses faith. The first point to be noticed is that the formula does not purport to express a convention -- let "p q" mean "it is not the case that p is true and q is false" -- but rather what is actually meant by the term "God’’ in the thought of a believer. "Faith" in the sense in which it figures in Anselm’s proof does not mean faith in the existence of God: it means instead the content which expresses the nature of God. If faith meant no more than the former, Hartshorne would be correct in saying that the Argument is merely the deduction that God exists from the premise "God exists." But Anselm’s premise is not that "God exists" but, rather, that the term "God" or ‘‘that than which nothing greater . . . ." when properly understood, leads to a contradiction when one and the same person claims to understand this term without being bound to acknowledge the impossibility of the divine non-existence. Anselm’s proof is dependent upon "faith," but not faith in the divine existence; instead, "faith’’ means the content of the idea of "God."2

I am inclined to think that Hartshorne’s reinterpretation of perfection in terms of the self-surpassing individual has the merit of overcoming the static connotation invariably associated with the perfect and of recovering the ideas of life and spirit within the divine nature. It is, however, necessary to see how this concept is related to "faith.’’ It seems to me that Hartshorne has three options: (1) he may claim that his conception is what Christian believers do mean by "God," or (2) what they would mean if they properly understood their faith, or (3) what "God" must mean if the argument is to succeed. I believe that Hartshorne can bypass option (1) because the issue turns in the end on a matter of principle, but I see him as committed to (2) and (3), but then he must maintain that the meaning in both these cases is coincident.

There need be no problem in calling attention to the dependence of the argument on faith and to the fact that faith is the source of at least one of its premises. This dependence of itself does not serve to show that there is no logical transition in the argument. The validity of the argument must turn on how the premise is understood and not upon its source. In this connection I would say that Anselm’s "discovery" -- he may have made more than one -- is that the divine nonexistence is not a real possibility because it contradicts the meaning of "God" properly understood.


I. The Logic of Perfection, p. 54.

2. "Anselm’s Discovery: A Re-examination of the Ontological Proof for Gods Existence," The Journal of Religion 47, no. 4 (1967): 365.

Response by Charles Hartshorne

In reading a book or essay I often write y for ‘‘yes" in margins. No’s are less frequent. I put only one no in reading Smith’s essay, and I suspect it is less a disagreement than a slight forgetting on his part, when he wrote that I assign God "to the domain of formal, pure, or necessary truth rather than to the side of fact." I assign only the bare, abstract truth that God exists to the nonfactual domain, but not the full concreteness of the divine life, which I call the divine "actuality." I never say that God is purely formal or necessary, only that the divine existence is so. The divine actuality, which is how, or in what state or states, the essence is actualized, is definitely contingent or factual. This is an aspect of what I call "dual transcendence.’’ Smith says that my principle of "modal coincidence’’ (to be possible is to be possible for God, to be actual is to be actual for God) is addressed to the problem of the factual side, but that my ontological argument must appeal to modal logic as expressive of real possibility and necessity, and that logicians express doubts about these. I agree that my argument must so appeal, and that many logicians do express such doubts.

Against such logicians I appeal to a number of considerations. One is that my theory of real possibility and necessity has much in common with views of three great logicians of the past, Aristotle, Peirce, and Whitehead. With them I take seriously the apparent asymmetry of becoming, time’s arrow, according to which the past is (in Peirce’s words) ‘‘the sum of accomplished facts,’’ of definite particulars, whereas the future is exclusively constituted by real Thirds, that is, not fully particularized generals, which will be somehow particularized as the future becomes past but are not particularized in advance or eternally. The eternal necessity of some actualization of divinity, God’s existence, is what all possible futures have in common. It is infinitely less particular than any given real possibility or actuality.

This theory is between the extremes, sheer denial of Objective or ontological modality and the Leibnizian type of "possible worlds" theory which seems to have some vogue among recent logicians. Aristotle, Peirce, and Whitehead do not use the concept of possible world; I also avoid it, except sometimes as shorthand. There are always contrasting possible future states of the actual world. Thus there may now be a real possibility of Reagan’s being reelected and also a possibility of his not running for a second term or of being defeated. Some logicians take this tack. According to it there may be possible states of given individuals but there are no merely possible, yet fully definite, particulars. Unqualified definiteness, particularity, and actuality are coincident. And Peirce says flatly, "It is the past which is actual." Real possibility is real futurity. Becoming is creation of definiteness, new Firsts, Seconds, and Thirds added to the already accomplished ones.

I give several criteria for the contingent as distinguished from the strictly necessary. Any positive conception whose instantiation restricts that of some other positive conceptions (for example as red-here-now restricts the occurrence of green-here-now) refers to a contingent aspect of reality. Incompatibility, not merely between P and not-P. but between P and Q, as two equally affirmative propositions, indicates contingency. I argue that the bare existence of God restricts no positive possibility whatever. God could coexist with anything else capable of existing. This is why theism cannot be observationally falsified, or in the proper sense empirically tested. As Popper says, observation is always of the presence, never of the mere absence, of something positive. Even black holes are not mere nothings.

George L. Goodman has discussed the relation of my ontological argument to formal logic in his book on that subject. William Lucas has proposed, in a dissertation completed in Austin, a formal system in which the problems can be discussed.

It still seems to me that "meaning-postulates" (Carnap’s phrase) must be given for general ideas beyond the recognized logical constants; one cannot define theism by these constants alone. But the ideas are no less universal in the metaphysical sense than are the logical constants. Whatever is implied by the mere meanings of Plato’s idea of Good, or of value, and "better than," or of Whitehead’s creativity, the ideas of dependence, independence, and still others (coherently combined) is metaphysically necessary and eternally true. Not only the divine existence but the actuality of some nondivine existence, is thus ontologically necessary. The bare idea of world-as-such is as ultimate as that of God, the ground being the equal tolerance for the positive coexistence of whatever you please. A worldless God is on the same footing of absurdity as a Godless world. But our actual world is purely factual, with no necessity requiring it.

I largely agree with what Smith says about the religious meaning of the term "God." I argue that, the world over, there has been some idea of a transcendent reality appropriate to the first ‘‘great" commandment, "Love the Lord thy God with all thy mind and heart and strength." I take this to imply that the transcendent reality must be somehow all-inclusive; that it must itself love all other realities and be related to them analogously to the way a human soul, that is, a personally ordered society of human experiences, is related to its body; also as a parent is related to its child, or as a ruler is to the subjects. Divine right of kings has always been fictitious, but deity is indeed, by eternal right, Lord or ruler of all -- not, however, by making decisions determining all that happens, leaving nothing for others to decide. Whoever seeks to do that is a tyrant, not a proper ruler. The word "omnipotent" stands for a human mistake, among the greatest of all such mistakes. It does not describe the one we are to love with all our being. In this contention process theism is not merely Western; Sri Jiva Goswami of Bengal was such a theist; so was Iqbal, the Islamic thinker and poet of Pakistan. But they were not classical theists of the medieval Western type. They were closer to the theologians of the Socinian sect in the seventeenth century.

Since the publication of my two books focusing on the ontological argument, the most important things I have done have been little noticed. One is to sketch a logic of "ultimate contrasts," with some analogy to the Hegelian dialectic; the other is to work out a new form of theistic arguments, some six of them, the ontological being only one, but the others being equally a priori. Both topics are the subjects of chapters in my Creative Synthesis and Philosophic Method. The scarcity of responses to these chapters illustrates my contention that it is possible to publish important ideas, reasonably well stated, virtually without attracting the attention of those concerned with the subject. The world, including the scholarly world, is always busy, always prejudiced. Various factors, in addition to good writing, must combine to gain its attention.

The real argument for God is the total context in which about six arguments, encapsulated in what I call the global argument, need to be placed. The chief contribution of the ontological argument is to make explicit the logical status of the theistic question, its transcendence of observational falsification. For the believer "the heavens declare the glory of God," but not even for the atheist can the heavens, or any observed realities, declare the divine nonexistence. To suppose that they could violates the logic of the idea of deity. Nontheism must argue on logical grounds, using "logical" here more broadly than some logicians would, not on observational grounds. This was Anselm’s discovery, but Aristotle already knew it. "With eternal things, to be possible and to be are the same."

Formalization of a theistic argument is only as convincing as the intuitions supporting its premises. But formalization helps to articulate the extra-logical premises, the intuitive content, of a belief. Anselm’s premises (as I revise his procedure) were two: there is a coherent idea of God, as all-surpassing, rivalry-excluding; this idea entails its own actualization, not how, or in what concrete actuality, it is actualized, but that it is somehow actualized, in some concrete form. The being "somehow actualized’’ of an essence or property is existence, as I use the word; the how or in what it is actualized is actuality. The latter is always contingent, even in the divine case; the former is contingent except in the divine case and whatever is implied by the bare idea of nondivine reality, some world or other.